Effects of P2Y₁₂ receptor antagonists beyond platelet inhibition – comparison of ticagrelor with thienopyridines

Abstract

The effect and clinical benefit of P2Y₁₂ receptor antagonists may not be limited to platelet inhibition and the prevention of arterial thrombus formation. Potential additional effects include reduction of the pro‐inflammatory role of activated platelets and effects related to P2Y₁₂ receptor inhibition on other cells apart from platelets. P2Y₁₂ receptor antagonists, thienopyridines and ticagrelor, differ in their mode of action being prodrugs instead of direct acting and irreversibly instead of reversibly binding to P2Y₁₂. These key differences may provide different potential when it comes to additional effects. In addition to P2Y₁₂ receptor blockade, ticagrelor is unique in having the only well‐documented additional target of inhibition, the equilibrative nucleoside transporter 1. The current review will address the effects of P2Y₁₂ receptor antagonists beyond platelets and the protection against arterial thrombosis. The discussion will include the potential for thienopyridines and ticagrelor to mediate anti‐inflammatory effects, to conserve vascular function, to affect atherosclerosis, to provide cardioprotection and to induce dyspnea.

Abbreviations

ACS

acute coronary syndrome

AngII

angiotensin II

ApoE

apoprotein E

CABG

coronary artery bypass grafting

CAD

coronary artery disease

CD39

ectonucleoside triphosphate diphosphohydrolase 1

CD62P

P‐selectin

CD73

ecto‐5′‐nucleotidase

CHD

coronary heart disease

EC

endothelial cell

ENT1

equilibrative nucleoside transporter 1

FBF

forearm blood flow

FMD

flow‐mediated dilatation

hsCRP

high sensitive C reactive protein

NSTEMI

non ST‐elevation myocardial infarction

PCI

percutaneous coronary intervention

ROS

reactive oxygen species

STEMI

ST‐elevation myocardial infarction

VSMC

vascular smooth muscle cells

Tables of Links

TARGETS

GPCRs a

Adenosine A2B receptor

P2Y12 receptors

Enzymes b

Akt

CD73, ecto‐5′‐nucleotidase

COX2

eNOS, endothelial NOS

ENT 1, equilibrative nucleoside transporter 1, SLC29A1

PKA

LIGANDS

Adenosine

Aspirin

cAMP

Cangrelor

Clopidogrel

IL‐6

MRS1754

NO

Prasugrel

Ticagrelor

Ticlopidine

TNF‐α

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (^{a b}Alexander et al., 2015a, 2015b).

Introduction

The P2Y₁₂ receptor is a key player in primary haemostasis and in arterial thrombosis and is an established target for antithrombotic drugs including ticlopidine, clopidogrel, prasugrel and ticagrelor.

Ticlopidine, clopidogrel and prasugrel are thienopyridines, which are all prodrugs that require hepatic enzyme‐dependent conversion into an active metabolite that irreversibly bind to P2Y₁₂ receptors. Ticlopidine was introduced in 1980, 21 years before its target receptor was cloned and characterized (Hollopeter et al., 2001). Ticlopidine has been gradually replaced by clopidogrel, which was launched in 1997. Clopidogrel's pharmacodynamic effect varies from 0% to 100% inhibition of ADP‐induced platelet aggregation (Järemo et al., 2002; Aleil et al., 2005). Prasugrel, which was introduced in 2009, has a less variable platelet inhibition driven by a more rapid, complete and consistent generation of its active metabolite, compared with clopidogrel (Wallentin et al., 2008). Ticagrelor, first approved in 2010, is the first member of a new chemical class, cyclopentyltriazolopyrimidines. Ticagrelor is directly acting, reversibly binding to the P2Y₁₂ receptor (van Giezen et al., 2009) and has an additional mode of action, not present for the thienopyridines, as ticagrelor also inhibits cellular adenosine uptake via the equilibrative nucleoside transporter 1 (ENT1) (Armstrong et al., 2014; Cattaneo et al.,2014).

Clopidogrel provides superior clinical benefit compared with aspirin (Gent et al., 1996) and compared with placebo when given in addition to aspirin in acute coronary syndrome (ACS) patients (Yusuf et al., 2001). Both prasugrel and ticagrelor provide superior clinical benefit compared with clopidogrel when given on top of aspirin in ACS patients (Wiviott et al., 2007; Wallentin et al., 2009). In addition, ticagrelor have shown superior clinical benefit compared with placebo when given on top of aspirin in patients with a prior myocardial infarction (Bonaca et al., 2015). True for all antiplatelet agents is that they do increase the risk of bleeding and that the clinical benefit in the studies listed previously is accompanied with an increased bleeding risk including life‐threatening bleeds. Thus, careful consideration of the benefit–risk ratio for each patient is needed to optimize the anti‐platelet therapy in terms of drugs to use and length of treatment on an individual level.

In addition to these already launched P2Y₁₂ receptor antagonists, there are two other non‐thienopyridine P2Y₁₂ receptor antagonists that have been tested in patients. Cangrelor, an ATP analogue, is a direct acting and reversibly binding P2Y₁₂ receptor antagonist. However, unlike ticagrelor, cangrelor is intended for acute treatment only, as the compound is short acting (t1/2 3–6 min) and not orally administered (Lhermusier et al., 2015). Acute treatment with cangrelor is superior to standard clopidogrel treatment when given as a bolus and infusion to patients undergoing urgent or elective percutaneous coronary interventions (PCI) (Bhatt et al., 2013). Cangrelor has just recently been approved as an adjunct to PCI. Finally, elinogrel is also a direct acting reversibly binding P2Y₁₂ receptor antagonist, but belongs to a different chemical class from ticagrelor and cangrelor. Elinogrel was intended both for p.o. and i.v. administration, but development was stopped in Phase II clinical trials.

This review will focus on the effects of ticagrelor and the thienopyridines and will summarize the evidence for effects of P2Y₁₂ receptor antagonists beyond platelet inhibition in the cardiovascular system.

Drug exposure

The pharmacodynamic effect of P2Y₁₂ receptor antagonists, on or off target, is related to the exposure of the active drug. The active metabolites of the thienopyridines can only be detected in the systemic circulation for about 4 h following intake of the prodrug and then only at levels below their in vitro IC₅₀ values, determined as the concentration providing 50% inhibition of ADP‐induced platelet aggregation (Figure 1B). One can therefore argue that the thienopyridines have limited or no systemic exposure, and one may hypothesize that the platelet inhibition by thienopyridines mainly takes place in the hepatic circulation where the active metabolites are formed and therefore likely present at higher concentrations. Because of their irreversible mode of action, however, thienopyridines can overcome the shorter duration of action of their active metabolite so that they still can provide platelet inhibition over 24 h.

Figure 1.

Illustration of plasma exposure of active drug over 24 h following ticagrelor 90 mg twice daily (A) and clopidogrel 75 mg once daily or prasugrel 10 mg once daily (B). In A, ticagrelor (dotted red line), ARC124910XX (dashed red line) and the sum of ticagrelor and AR‐C124910XX (solid red line). In B, clopidogrel active metabolite (AM) (purple line) and prasugrel‐AM (green line). Illustration is based on published patient exposure data (Wallentin et al., 2008; Storey et al., 2007). Dotted black lines represent IC50 values, using ADP‐induced light transmission aggregometry in human PRP in vitro. In A, IC50 is the mean of ticagrelor and AR‐C124910XX data presented in Table 1. In B, clopidogrel‐AM and prasugrel‐AM IC50 data is from Sugidachi et al., 2007.

In contrast to thienopyridines, ticagrelor does not require hepatic enzyme‐dependent conversion into an active metabolite. That said ticagrelor has a major circulating metabolite, AR‐C124910XX, with plasma exposure of 30–40% (Storey et al., 2007) and similar potency versus P2Y₁₂ as the parent ticagrelor (Table 1). Ticagrelor is given twice a day, which provides a biphasic increase in the plasma concentration of ticagrelor and AR‐C124910XX over 24 h (Figure 1A). Given that both ticagrelor and its metabolite are active and reversibly binding to the platelet P2Y₁₂ receptor, the main pharmacodynamic effect, platelet inhibition, is directly related to the combined exposure of ticagrelor and AR‐C124910XX in the systemic circulation (Husted et al., 2012). As predicted from the in vitro potency data, a high and continuous platelet inhibition is achieved in patients as the plasma exposure of ticagrelor is well above its in vitro IC₅₀ throughout 24 h (Figure 1). The presence of an active drug in the systemic circulation throughout 24 h provides what can be referred to as a systemic potential.

Table 1.

The mean concentrations that gave half‐maximum inhibition (IC₅₀) of ticagrelor and AR‐C124910XX with regards to receptor binding, receptor signalling and ADP‐induced platelet aggregation in whole blood and platelet‐rich plasma (PRP) is listed together with the binding affinity (K_i) at platelet P2Y₁₂ receptors

In vitro inhibition of human P2Y12 receptor radioligand binding and 2MeSADP‐induced receptor signalling
	Receptor binding, IC50 μM, (n)	Ki, nM, (n)	Receptor signalling, IC50 μM, (n)
Ticagrelor	0.01 (63)	2.0 (8)	0.07 (98)
AR‐C124910XX	0.06 (2)	2.5 (2)	0.05 (2)

In vitro inhibition of ADP‐induced platelet aggregation in human whole blood and PRP
	Whole blood, IC50 μM, (n)	PRP, IC50 μM, (n)
Ticagrelor	0.24 (4)	0.40 (8)
AR‐C124910XX	0.17 (4)	0.13 (8)

For receptor binding the displacement of [125I]AZ11931285, a radio‐ligand that binds to the same site on the receptor was used, and for receptor signalling, the inhibition of 2MeSADP‐induced P2Y₁₂ receptor signalling using a 35S‐GTPγS assay was evaluated. Both assays used membranes from human P2Y₁₂ receptor transfected cells. ADP‐induced platelet aggregation was evaluated in whole blood with impedance aggregometry using Multiplate (Dynabyte, Munich, Germany) and in PRP with light transmission aggregometry using PAP8 (BioData Corporation, PA, USA). The data is unpublished except for the Ki data for ticagrelor (Springthorpe et al., 2007). Overall data indicate similar potency at P2Y₁₂ receptors for ticagrelor and AR‐C124910XX.

This systemic potential means that ticagrelor apart from platelet P2Y₁₂ receptors should be able to reach additional targets (non‐platelet P2Y₁₂ receptors and the equilibrative nucleoside transporter 1, ENT1 for ticagrelor) expressed systemically.

ENT1 inhibition by ticagrelor

Adenosine is formed from ATP and ADP released locally at sites of ischaemia, tissue damage and inflammation. The t1/2 of adenosine in the circulation is a matter of seconds due to its rapid cell uptake and intracellular metabolism. Ticagrelor inhibits the ENT1 transporter and thereby reduces the cellular uptake of adenosine resulting in its prolonged local t1/2 and extracellular presence (Nylander et al., 2013; Armstrong et al., 2014). Ticagrelor has been shown to increase the plasma levels of adenosine in ACS patients (Bonello et al., 2013) and to augment adenosine‐induced physiological responses, including increases in coronary blood flow (Alexopoulos et al., 2013; Wittfeldt et al., 2013), platelet inhibition (Nylander et al., 2013), neutrophil migration (Alsharif et al., 2015) and the sensation of dyspnea (Wittfeldt et al., 2013). The additional mode of action, ENT1 inhibition and its potential clinical relevance has recently been reviewed (Cattaneo et al., 2014). Importantly, and unlike P2Y₁₂ receptors, only ticagrelor and not AR‐C124910XX significantly inhibits ENT1 (Armstrong et al., 2014). In addition to ENT1 inhibition, ticagrelor induced ATP release from human erythrocytes in vitro (Öhman et al., 2012); it remains unknown whether this effect of ticagrelor on erythrocytes occurs also in vivo. If present, both mechanisms would work together to further enhance the local concentration of adenosine as ATP is rapidly converted into AMP via the ectonucleoside triphosphate diphosphohydrolase 1 (CD39), with subsequent conversion into adenosine via the ecto‐5′‐nucleotidase (CD73) (Cattaneo et al., 2014).

Juvenile platelets

Patients with increased platelet turnover have higher levels of immature platelets, and an increase in the immature platelet count has been found to be associated with worse cardiovascular outcome (Cesari et al., 2013; Ibrahim et al., 2014) supporting the hypothesis that these juvenile platelets are more pro‐thrombotic. A continuous systemic exposure of an active compound, as true for ticagrelor, should theoretically inhibit newly formed, juvenile platelets equally well as the old platelets. This should not be the case for the thienopyridines as the platelets formed after elimination of the active metabolites from the circulation will not be inhibited until the next dose of prodrug on the next day. In support, high immature platelets count has been found associated with lower anti‐platelet response to clopidogrel (Ibrahim et al., 2012). A rat study has shown that the recovery of platelet function after ticagrelor administration differs from that after clopidogrel treatment (Kuijpers et al., 2011). Following ticagrelor administration, all platelets (old and newly formed) were inhibited, and their function gradually recovered over time in parallel with drug elimination. Following clopidogrel treatment, there was a gradually emerging subpopulation of uninhibited platelets, probably juvenile platelets. This difference in the platelet inhibition properties is masked by conventional platelet aggregation tests, which evaluate the net response of the total platelet population, but was revealed by thrombus formation measurement under flow where the juvenile un‐inhibited platelets, formed at later time points after clopidogrel treatment, appeared to support thrombus formation (Kuijpers et al., 2011). In another study a mixture of uninhibited and prasugrel‐inhibited platelets was studied to model the role of juvenile platelets formed each day. Twenty percent prasugrel‐inhibited plus 80% uninhibited platelets were used to mimic diabetics that have a high platelet turnover with an approximate platelet lifespan of 5 days (thus about 20% new platelets should be formed each day). By imaging of ADP‐induced aggregates of these mixtures, the authors showed that uninhibited platelets supported a core of aggregated platelets, which was surrounded by prasugrel‐inhibited platelets (Hoefer et al., 2015; Gurbel and Tantry, 2015). Another recent study explored platelet function after discontinuation of prasugrel in ACS patients. They found that prasugrel discontinuation resulted in the formation of an emerging subpopulation of ADP‐responsive juvenile platelets, which contributed to platelet aggregation and thrombus formation under flow (Baaten et al., 2015). Finally, a clinical study explored platelet function in ACS patients on ticagrelor or prasugrel treatment (Bernlochner et al., 2015). They found that the number of juvenile platelets (immature platelet count) correlated with platelet aggregation in patients treated with prasugrel, but not in those receiving ticagrelor, indicating that juvenile platelets contribute to increased platelet aggregation in prasugrel‐treated, but not with ticagrelor‐treated patients. Flow cytometry experiments confirmed that the juvenile platelets were less inhibited in prasugrel, compared with ticagrelor‐treated patients especially late after drug intake (1 h before next dose).

Non‐platelet P2Y ₁₂ receptors

Historically, P2Y₁₂ receptors have been seen as platelet‐specific. More recently, it has become clear that many cells express P2Y₁₂ receptors including endothelial cells (EC) (Simon et al., 2002; Shanker et al., 2006), vascular smooth muscle cells (VSMC) (Wihlborg et al., 2004; Satonaka et al., 2015), dendritic cells (Ben Addi et al., 2010), leukocytes (Wang et al., 2004; Diehl et al., 2010) and neurons (Kawaguchi et al., 2015). The roles of these non‐platelet P2Y₁₂ receptors are poorly explored, but data are emerging (Gachet, 2012).

Activation of P2Y₁₂ receptors expressed on EC decreased intracellular cAMP content (Simon et al., 2002). EC cAMP content, via activation of PKA, was responsible for endothelial barrier function (Lum et al., 1999), which was disturbed under pathophysiological conditions including increased inflammation (Aslam et al., 2014) or hypoxia/reoxygenation (Aslam et al., 2013). Disruption of endothelial barrier function was attenuated by elevation of EC cAMP levels (Aslam et al., 2013, 2014) which might occur following blockade of P2Y₁₂ receptors (Simon et al., 2002).

The VSMC P2Y₁₂ receptors have been shown to induce vasoconstriction when activated by ADP (Wihlborg et al., 2004). Ticagrelor inhibited this ADP‐induced VSMC vasoconstriction in vitro whereas in vivo thienopyridine treatment had no effect, most likely because of no or low presence of the active metabolites at the site of the target (Högberg et al., 2010; Grzesk et al., 2012) (Figure 1) The active metabolite of prasugrel, when added in vitro, inhibited the P2Y₁₂ receptor agonist 2‐methylthio‐ADP‐induced VSMC IL‐6 expression and mitogenesis (Rauch et al., 2010). The clinical relevance of non‐platelet P2Y₁₂ receptors and their inhibition, including VSMC P2Y₁₂, is unknown.

A hypothesis has been proposed that non‐platelet, potentially neuronal P2Y₁₂ receptors, could be the mediators of drug‐induced dyspnea observed in patients (Cattaneo and Faioni, 2012). This hypothesis is supported by the observation that all direct acting and reversibly binding P2Y₁₂ receptor antagonists evaluated in patients (ticagrelor, cangrelor and elinogrel) increased the incidence of dyspnea. In the PLATO study, the incidence of dyspnea was 13.8% versus 7.8% in ticagrelor‐treated and clopidogrel‐treated patients and in the PEGASUS study, 18.9%, 15.8% and 6.4% experienced dyspnea when 90 mg ticagrelor, 60 mg ticagrelor or placebo was given on top of aspirin (Wallentin et al., 2009; Bonaca et al., 2015). In the INNOVATE‐PCI study, elinogrel induced 12.3% dyspnea versus 3.8% for clopidogrel (Welsh et al., 2012). In studies with cangrelor, the incidence of dyspnea is much lower, probably because of the short duration of treatment, but still highly significant (Bhatt et al., 2009; Bhatt et al., 2013; Harrington et al., 2009). The characteristics of ticagrelor‐induced dyspnea in the PLATO study has been published, and the authors concluded that it was usually mild or moderate in intensity and not associated with differences in efficacy or safety outcomes in the study (Storey et al., 2011). As adenosine is known to induce dyspnea (Burki et al., 2005), an alternative hypothesis to explain ticagrelor‐induced dyspnea is increased extracellular adenosine levels via ENT1 inhibition. Indeed, as mentioned, ticagrelor enhanced adenosine‐induced dyspnea in healthy subjects (Wittfeldt et al., 2013). However, unlike ticagrelor, cangrelor and elinogrel do not inhibit cellular adenosine uptake (Armstrong et al., 2014), but still increased the incidence of dyspnea. Thus, the adenosine hypothesis does not explain what may be a class effect of these compounds (Cattaneo et al., 2014). As dyspnea is a sensation, it is hard to explore the mechanism responsible in animal models why dyspnea needs to be explored in patients that can express their level of breathlessness. Currently, a clinical trial is ongoing to explore the hypothesis that ticagrelor‐induced dyspnea is mediated via adenosine. In this study, ticagrelor patients experiencing dyspnea will be randomized to receive caffeine (an adenosine receptor antagonist) or placebo for 1 week, and the primary end point is change in visual analogue scale AUC for dyspnea (Lindholm et al., 2015).

In summary, the pharmacokinetic and pharmacodynamic properties of available P2Y₁₂ receptor antagonists have a number of key differences. This provides different potential to inhibit systemically expressed non‐platelet P2Y₁₂ receptors and off targets, such as inhibition of ENT1 in the case of ticagrelor. Furthermore, 24 h plasma exposure of active compound is required to provide continuous platelet inhibition of old and newly formed platelets alike.

Effects on inflammation

It is well known that platelets play a role in inflammatory responses (see Gros et al., 2015; Nording et al., 2015; Thomas and Storey, 2015). Platelets store pro‐inflammatory cytokines that are released when platelets are activated. Also microRNAs, small non‐coding RNAs, such as micro RNA‐223, RNA‐126, RNA‐21, RNA‐24 and RNA‐197 are known to be present in platelets and are secreted in exosomes and/or microvesicles upon platelet activation (Hulsmans and Holvoet, 2013; Fuentes et al., 2015). These micro RNAs, once taken up by cells, can further trigger inflammatory processes and increase atherosclerosis and angiogenesis (see Fuentes et al., 2015). In addition, activated platelets translocate adhesion receptors, for example, CD62P (P‐selectin) and CD40L from their α granules to the plasma membrane. CD62P will bind to P‐selectin glycoprotein ligand‐1 on leukocytes and create platelet–leukocyte aggregates and trigger leukocyte activation. CD40L will bind to CD40 on a number of cells including EC and trigger their activation. As a result of leukocyte and EC activation, more pro‐inflammatory cytokines will be released, and more adhesion receptors expressed at the site of inflammation that further amplifies the inflammatory process (Figure 2). Thus, inhibition of platelet activation will attenuate the pro‐inflammatory capacity of platelets. All anti‐platelet drugs should have this effect to different extent depending on the importance of the target and the level of target inhibition.

Figure 2.

Illustration of the role of platelets in amplification of inflammation via interactions with leukocytes and endothelial cells and the release of cytokines from all cells involved. Platelets store pro‐inflammatory cytokines that are released when platelets are activated. In addition, activated platelets translocate adhesion receptors, for example, CD62P (P‐selectin) and CD40L from their α granules to the plasma membrane. CD62P will bind to P‐selectin glycoprotein ligand‐1 on leukocytes and create platelet–leukocyte aggregates and trigger leukocyte activation. CD40L will bind to CD40 on a number of cells including endothelial cells (EC) and trigger their activation. As a result of leukocyte and EC activation, more pro‐inflammatory cytokines will be released, and more adhesion receptors expressed at the site of inflammation that further amplifies the inflammatory process. This pro‐inflammatory role of platelets can have important implications in inflammatory processes including atherosclerosis.

Thienopyridines

In line with the reasoning outlined above , inhibition of P2Y₁₂ receptors in vitro by clopidiogrel's active metabolite reduced P‐selectin expression, platelet‐polymorphonuclear leukocyte adhesion and production of reactive oxygen species (ROS) by polymorphonuclear leukocytes (Evangelista et al., 2005). In rabbits, vascular inflammation following myocardial ischaemia/reperfusion was reduced by clopidogrel associated with preserved NO synthase (NOS) expression following clopidogrel treatment, compared with placebo (Molero et al., 2005).

The active metabolite of prasugrel yielded a concentration‐dependent inhibition of platelet aggregation, soluble CD40L (sCD40L) release and platelet–leukocyte co‐aggregate formation in human blood in vitro (Judge et al., 2008). The latter finding was confirmed by others in human blood samples (Frelinger et al., 2007) and extended to the inhibition of agonist‐stimulated platelet–monocyte adhesion in blood from prasugrel treated mice (Totani et al., 2012) (see Schrottmaier et al., 2015, for the platelet–leukocyte/monocyte interaction). Apart from inhibiting platelet–leukocyte and platelet–monocyte adhesions, prasugrel reduced TNF‐α synthesis and increased NO metabolites in endotoxin‐treated mice in vivo (Totani et al., 2012). Interestingly, the anti‐inflammatory actions of the active metabolite of prasugrel was likely derived from direct targeting of neutrophils isolated from human blood and was P2Y₁₂ receptor independent (Liverani et al., 2013). This has been recently reviewed by Thomas and Storey (2015.

Switching aspirin (75 mg·day⁻¹) to clopidogrel (75mg·day⁻¹) reduced the concentrations of some inflammatory markers (in particular high sensitive C reactive protein (hsCRP), IL‐6 and CD40L) in diabetic patients with remaining high platelet reactivity (Rosiak et al., 2013). In contrast, in patients with coronary heart disease (CHD), no between‐group differences in circulating markers of inflammation after 1 year treatment with clopidogrel (75 mg·day⁻¹) compared with aspirin (160 mg·day⁻¹) was evident, but both treatments reduced the levels of TNF‐α (Solheim et al., 2006). Adding clopidogrel on top of aspirin further reduced platelet CD40L expression (Quinn et al., 2004) and reduced sCD40L (Azar et al., 2006). Clopidogrel abolished the influence of leukocytes on platelets (Diehl et al., 2010) and decreased the elevated platelet–monocyte and platelet–neutrophil conjugates in the blood of ACS patients and prevented their agonist‐triggered formation ex vivo (Xiao and Théroux, 2004), while such an effect was not seen in patients with non ST‐elevation myocardial infarction (NSTEMI) at 4 weeks of clopidogrel treatment (Palmerini et al., 2010). Also, the plasma level of hsCRP was reduced in patients on additional treatment with clopidogrel, but not in patients on placebo (Heitzer et al., 2006; Chen et al., 2006). In ACS patients undergoing PCI, hsCRP increased significantly at day 5 in clopidogrel naïve patients, but was reduced in patients on clopidogrel (Vivekananthan et al., 2004). Similarly, in low clopidogrel responders, hsCRP increased 7 days post PCI while in high responders, hsCRP remained within the reference intervals (Malek et al., 2007). In patients receiving long‐term clopidogrel therapy compared with the clopidogrel‐naive group, lower levels of selected inflammation markers (IL, TNF‐α and hsCRP) were achieved (Antonino et al., 2009).

In support of the effects of clopidogrel, in patients undergoing PCI, the plasma levels of sCD40L, hsCRP and IL‐6 after 6 month were higher in patients with clopidogrel resistance than in those with normal clopidogrel responsiveness (Ge et al., 2012; see also Woo et al., 2014). Clopidogrel withdrawal was associated with an increase in platelet and inflammatory biomarkers in diabetic patients, again supporting anti‐inflammatory affects coupled to P2Y₁₂ receptor antagonism (Angiolillo et al., 2006). While the vast majority of studies with clopidogrel have demonstrated an anti‐inflammatory effect, a few studies have failed to show a reduction in hsCRP (Table 2). Clopidogrel therapy also had no consistent effect on EC function, arterial stiffness and oxidative stress markers (Ostad et al., 2011), or the number of circulating progenitor cells (Ramadan et al., 2014); see Husted et al., 2010; Willoughby et al., 2014). Angiogenic cytokines influence vessel injury, and platelets represent a disposable circulating pool of angiogenic molecules. In 28 healthy male volunteers treated for 7 days with clopidogrel, the level of circulating angiogenic factor (VEGF, placenta growth factor and stromal cell‐derived factor‐1) remained unchanged (Smadja et al., 2012).

Table 2.

Studies on clopidogrel describing its direct platelets effects, effects on platelet‐leukocyte interaction, effects on inflammation or vascular reactivity

Clopidogrel	Patients	Aspirin	Platelet (sCD40L, CD62)	Platelet‐leukocyte interaction	Inflammation (hsCRP, TNF, IL)	Oxidative stress	Vascular reactivity (FMD, FBF)	Reference
24 h 5 days	ACS PCI (n = 833)	Yes	NA	NA	unchanged reduced	NA	NA	Vivekananthan et al., 2004
24 h	ACS (n = 23)	Yes	reduced	reduced	NA	NA	NA	Xiao and Théroux, 2004
5 days	PCI (n = 79)	Yes	reduced	NA	unchanged	NA	NA	Quinn et al., 2004
24 h	ACS (n = 103)	Yes	unchanged	NA	unchanged	NA	NA	Montalescot et al., 2006
24 h	PCI (n = 60)	Yes	NA	NA	TNF reduced CRP unchanged	NA	NA	Gurbel et al., 2006
1–4 weeks	ACS (n = 115)	Yes	NA	NA	reduced	NA	NA	Chen et al., 2006
5 weeks	CAD (n = 103)	Yes	reduced	NA	reduced	NA	increased	Heitzer et al., 2006
8 weeks	CAD (n = 73)	Yes	reduced	NA	unchanged	NA	NA	Azar et al., 2006
12 months withdrawal	Diabetes PCI (n = 54)	Yes	increased (upon withdrawal)	NA	Increased (upon withdrawal)	NA	NA	Angiolillo et al., 2006
12 month	CHD (n = 101)	No	unchanged	NA	TNF reduced CRP unchanged	NA	NA	Solheim et al., 2006
responder versus non‐responder 7 days	STEMI, PCI (n = 34)	Yes	NA	NA	decreased	NA	NA	Malek et al., 2007
22 hours	CAD (n = 58)	Yes	NA	NA	NA	NA	increased	Warnholtz et al., 2008
Naive versus 6 month	PCI (n = 110)	Yes	NA	NA	reduced	NA	NA	Antonino et al., 2009
1–4 weeks	STEMI PCI (n = 54)	Yes	NA	unchanged	reduced	NA	NA	Palmerini et al., 2010
4 weeks	NTSE‐ACS (n = 990)	Yes	unchanged	NA	unchanged	unchanged	NA	Husted et al., 2010
28 days	CAD (n = 120)	Yes	NA	NA	unchanged	unchanged	unchanged	Ostad et al., 2011
6 months responder versus non‐responder	PCI (n = 352)	Yes	reduced	NA	reduced	NA	NA	Ge et al., 2012
8 weeks	CAD, diabetes (n = 234)	switching to clopidogrel	reduced	NA	reduced	NA	NA	Rosiak et al., 2013
6 weeks	CAD (n = 41)	Yes	reduced	NA	unchanged	unchanged	unchanged	Ramadan et al., 2014
3 month	CAD (n = 35)	Yes	NA	NA	NA	unchanged	increased	Willoughby et al., 2014
498 days	CAD (n = 91)	Yes	reduced	NA	reduced	NA	increased	Woo et al., 2014
4 weeks Responder vs. Non‐responder	CAD, PCI (n = 108)	Yes	NA	NA	NA	Na	increasd	Siasos et al., 2015

For abbreviations, see List of abbreviations. NA: not assessed. For review see also (Muhlestein, 2010; Steinhubl et al., 2007; Yeh and Khan, 2006).

In patients with sickle cell disease, prasugrel reduced the number of platelets adhering to monocytes and neutrophils (Jakubowski et al., 2014). While the in vivo levels of platelet–monocyte aggregates did not differ between patients receiving clopidogrel or prasugrel treatment, the levels were significantly higher in blood from patients treated with clopidogrel as compared with prasugrel following in vitro platelet activation (Gremmel et al., 2013). However, such difference was not seen after 1 month treatment in another study in patients with stable CHD (Braun et al., 2008). An anti‐inflammatory response of prasugrel was also tested in asthmatic patients, randomly and blindly allocated to prasugrel or placebo for 15 days. In this study, the provocative dose of mannitol causing a 15% drop in FEV₁, tended to increase from 142 to 187 mg, after prasugrel, but did not change after placebo treatment. These results suggest that P2Y₁₂ inhibition may reduce the bronchial inflammatory burden (Lussana et al., 2015). More recently, prasugrel, when given to healthy subjects, inhibited the increased release of pro‐inflammatory cytokines when isolated T‐cells from treated subjects was stimulated in the presence of isolated platelets ex vivo (Johnston et al., 2015).

Ticagrelor

In a mouse caecal ligation and puncture model of polymicrobial sepsis, ticagrelor reduced formation of neutrophil–platelet aggregates and lung damage driven by neutrophil recruitment (Rahman et al., 2014). In addition to anti‐inflammatory effects mediated by inhibition of platelet P2Y₁₂ receptors, adenosine is a potent modulator of inflammatory responses. The ability of ticagrelor to provide anti‐inflammatory effects via adenosine has been shown in vitro as ticagrelor enhanced adenosine‐induced neutrophil migration (Alsharif et al., 2015).

Clinically, a post hoc analysis of the PLATO trial revealed that ticagrelor compared with clopidogrel was associated with a lower morbidity and mortality related to pulmonary infection and sepsis (Storey et al., 2014). Thus, an anti‐inflammatory effect may have contributed to the overall mortality benefit seen in the PLATO trial. However, such a beneficial effect was not seen in the recent PEGASUS trial where numerically more deaths related to infection were seen with 90 mg relative to 60 mg ticagrelor and placebo (Bonaca et al., 2015). The reason for the discrepancy between the trials is unknown but may be related to the different patient populations studied. For instance, an anti‐inflammatory effect may have been especially important in patients with coronary artery bypass grafts (CABG), who are likely to be more prone to infection. In the PLATO trial, ticagrelor reduced mortality by 50% compared with clopidogrel in CABG patients (Held et al., 2011; Varenhorst et al., 2012]. In the more stable patients evaluated in the PEGASUS trial, less patients should have undergone CABG, which could in part explain the difference in infection or sepsis related mortality.

As ticagrelor confers higher and more consistent P2Y₁₂ receptor inhibition compared with clopidogrel, ticagrelor should be more effective with regards to any protective effects mediated by P2Y₁₂ receptor inhibition. Indeed, ticagrelor and clopidogrel reduced the release of the pro‐inflammatory cytokine TNF‐α, but only ticagrelor also significantly reduced IL‐6 in a model of human sepsis (Thomas et al., 2014). However, in the PLATO trial, there were no significant differences in inflammatory biomarkers hsCRP, IL‐6, myeloperoxidase or sCD40L in patients treated with ticagrelor versus clopidogrel (Husted et al., 2010). This discrepancy may be related to the time between biomarker sampling, occurrence of the potential inflammatory trigger and timing of last drug dose, which was not captured in the PLATO trial. Currently, the potential effects of ticagrelor on mortality is being assessed in patients with pneumonia (Ticagrelor in Severe Community Acquired Pneumonia (TCAP), ClinicalTrials.gov. Identifier: NCT01998399). The effects of P2Y₁₂ receptor antagonists on inflammation and immunity have been recently reviewed (Thomas and Storey, 2015).

In summary, the vast majority of studies show that inhibition of platelet P2Y₁₂ receptor reduced the amplification of inflammatory responses induced by activated platelets. The level of the anti‐inflammatory effect appears to depend on the level of P2Y₁₂ receptor inhibition although some P2Y₁₂ receptor‐independent effects have been described. Clinically, extensive P2Y₁₂ receptor blockade may result in reduced morbidity and mortality in patients with increased pro‐inflammatory burden.

Effects on vascular function

Thienopyridines

Clopidogrel released NO from cultured EC independently from its anti‐platelet action, and thienopyrimidinone congeners that were devoid of anti‐platelet action also stimulated coronary endothelium to release NO (Jakubowski et al., 2005). Similarly, clopidogrel caused relaxation of arterial tissues and influenced VSMC proliferation directly without hepatic biotransformation in vitro. The effect was independent of endothelium, β‐adrenoceptor and purinergic P2 receptors (Froldi et al., 2011).

In experimental disease models in vivo, endothelium‐dependent relaxation upon ACh stimulation was reduced in rats made hypertensive by angiotensin (Ang) II‐infusion, and clopidogrel treatment was effective in improving endothelial function. Clopidogrel also prevented vascular remodelling, shown by augmented media thickness in aortas from Ang II‐hypertensive rats (Giachini et al., 2014). Similarly, in chronic heart failure rats with reduced NO‐mediated vasorelaxation, clopidogrel treatment improved the vasorelaxation. The effect of clopidogrel was associated with an enhanced phosphorylation of Akt and endothelial NOS (Schäfer et al., 2011).

In patients with symptomatic coronary artery disease (CAD), clopidogrel improved endothelial NO bioavailability and diminished biomarkers of oxidant stress (Heitzer et al., 2006). A higher clopidogrel maintenance dose (150 mg·day⁻¹) was associated with a greater improvement of endothelial function compared with the standard 75 mg·day⁻¹ (Patti et al., 2011). Clopidogrel reduced endothelial injury (assessed as number of circulating EC) during PCI (Bonello et al., 2010) and improved vascular reactivity, either P2Y₁₂ receptor‐dependent (Hamilos et al., 2011; Siasos et al., 2015) or receptor‐independent (Warnholtz et al., 2008; Willoughby et al., 2014). However, such beneficial effect of clopidogrel on vascular reactivity was not seen in all clinical studies, and the beneficial effect of clopidogrel on flow‐mediated brachial artery dilation was not sustained after long‐term clopidogrel treatment (28 days) (Ostad et al., 2011) (Table 2).

Ticagrelor

Ticagrelor increased eNOS activation and COX2 protein expression in TNF‐α stimulated human aortic EC in vitro by a yet unknown mechanism (Reiner et al., 2013). Also, in vivo treatment with ticagrelor increased eNOS activation and both the activation and expression of COX2 in a rat model of ischaemia/reperfusion injury. These effects of ticagrelor were dependent on adenosine receptor activation with no effect of clopidogrel in spite of dosing to similar levels of P2Y₁₂ receptor blockade as ticagrelor (Nanhwan et al., 2014). In a prospective randomized clinical study, ticagrelor increased the number of circulating endothelial progenitor cell compared with clopidogrel in ACS patients (Bonello et al., 2015). Finally, in a non‐randomized study, ACS patients on ticagrelor treatment but not on clopidogrel or prasugrel treatment had improved peripheral endothelial function as evaluated with the EndoPAT system (Torngren et al., 2013). An ongoing study evaluates the acute and chronic effects of clopidogrel, prasugrel and ticagrelor on parameters of vascular and platelet function in ACS patients (Schnorbus et al., 2014).

In summary, thienopyrimidinones devoid of antiplatelet action increase NO production. Clopidogrel has been proven beneficial regarding EC/vascular function, part of it being P2Y₁₂ receptor‐independent. Similarly, ticagrelor increased eNOS but also COX2 expression and activity and, unlike clopidogrel and prasugrel, appears to improve peripheral EC function in ACS patients.

Effects on atherosclerosis

Many risk factors contribute to the accelerated development of atherosclerotic lesions; among them are hypertension, hypercholesterolemia and diabetes. In apoprotein E (ApoE) knockout mice, having a disturbed cholesterol metabolism, platelets adhered to the vascular endothelium of the carotid artery before the development of atherosclerotic lesions (Dong et al., 2000; Massberg et al., 2002). Platelet adhesion to the endothelium coincided with inflammatory gene expression and preceded atherosclerotic plaque invasion by leukocytes (Massberg et al., 2002). Once activated, platelets produce ROS via NADPH oxidases (Stokes et al., 2007; Martin‐Ventura et al., 2012), and ROS then further activate non‐adherent platelets and oxidized phospholipids including LDLs (oxLDL) (Freedman, 2008). OxLDL in turn stimulates platelet–monocyte aggregate formation, which promotes phenotypic changes in monocytes, monocyte extravasation and enhanced foam‐cell formation in vitro and in vivo. Inhibition of P2Y₁ and P2Y₁₂ receptors partly blocked oxLDL‐induced platelet–monocyte aggregation (Badrnya et al., 2014) (Figure 3).

Figure 3.

Effects of P2Y12 receptor activation on endothelial cell (EC), platelets and vascular smooth muscle cells (VSMC) function. For detailed explanation, see text. miR: micro RNA.

NO is an endogenous platelet inhibitor (Freedman, 2008), and platelet sensitivity towards NO is reduced in the presence of P2Y₁₂ receptor activation (Kirkby et al., 2013). As increased ROS formation by platelets may reduce NO availability, this should indirectly increase platelet activation and vascular tone. Activation of P2Y₁₂ receptors may occur early during atherosclerosis because circulating levels of the P2Y₁₂ receptor agonist ADP were elevated in patients with peripheral atherosclerotic disease especially of younger age (Jalkanen et al., 2015). In patients with peripheral atherosclerotic disease, inhibition of P2Y₁₂ receptors greatly increased platelet sensitivity towards NO (factor of 1000) thereby attenuating platelet activation and aggregation.

Studies using P2Y₁₂ knockout mice and bone marrow chimeric mice expressing either vessel wall or platelet P2Y₁₂ receptors indicate a contribution of the platelet P2Y₁₂ receptor to restenosis (Evans et al., 2009) and of the vessel wall and platelet P2Y₁₂ receptors to the development of atherosclerosis (Li et al., 2012; West et al., 2014).

Thienopyridines

In rabbits, with high‐fat diet‐induced atherosclerosis, treatment with clopidogrel attenuated plaque development and reduced the vascular wall and systemic inflammation (Li et al., 2007; Hadi et al., 2013). Two studies in ApoE knockout mice demonstrated beneficial effects of clopidogrel in attenuating atherosclerotic lesion formation (Afek et al., 2009; Takeda et al., 2012). However, other studies in ApoE knockout mice suggested that clopidogrel may not be effective in preventing early atherosclerosis (Schulz et al., 2008; West et al., 2014). Clopidogrel has also been shown to reduce neointimal formation (Waksman et al., 2008; Akbulut et al., 2004). Finally, an association between increased platelet reactivity during clopidogrel treatment and greater coronary artery atherosclerotic disease burden and plaque calcification has been described in patients undergoing PCI (Chirumamilla et al., 2012).

Ticagrelor

Supporting the role of platelets in the initiation of atherosclerosis, 8 weeks of ticagrelor treatment attenuated the initiation of atherosclerosis in hypercholesterolemic ApoE deficient mice (Schirmer et al., 2012). However, a similar study using the same model with only 4 weeks treatment showed no effect of ticagrelor or clopidogrel treatment on atherosclerosis development (West et al., 2014). The discrepancy between these studies can potentially be explained by drug dose, the time of initiation of treatment and the duration of the treatment.

Adenosine is a modulator of inflammatory responses and adenosine receptor deficient mice on an ApoE‐deficient background display alterations in atherosclerosis development (Jones et al., 2004; Reiss et al., 2004; Wang et al., 2009; Bingham et al., 2010; Koupenova et al., 2012; Teng et al., 2014) with more clear benefit seen in CD73 (ecto‐5′‐nucleotidase) and ApoE double knockout mice, which should have reduced adenosine production (Buchheiser et al., 2011). To address the potential impact of ticagrelor on established atherosclerosis, 20 weeks old ApoE knockout mice with established plaques were treated with ticagrelor for 25 weeks. In this study, ticagrelor induced a more stable plaque phenotype (thicker fibrous cap and reduced necrotic core) with a trend for a reduction in plaque size (Rusnak et al., 2014). Additionally, ticagrelor has been shown to reduce neointimal formation in a mouse carotid artery injury model (Patil et al., 2010) and reduced neointimal hyperplasia in a rabbit carotid anastomosis model (Sürer et al., 2014).

In summary, there is data from animal models supporting the hypothesis that thienopyridines and ticagrelor can affect the initiation and progression of atherosclerosis primarily via inhibition of P2Y₁₂ receptors. Patient imaging studies are needed to validate the animal findings and to compare thienopyridines and ticagrelor with regards to the development and progression of atherosclerosis.

Cardioprotection

Thienopyridines

In rabbits, pre‐treatment with clopidogrel for 2 or 3 days (Yang et al., 2013) reduced infarct size following ischaemia/reperfusion. Interestingly, the protective effect of clopidogrel could be attenuated with MRS1754, an adenosine A_2B receptor antagonist (Ye et al., 2013). When dosed to an equal P2Y₁₂ receptor blockade, ticagrelor but not clopidogrel reduced infarct size in ischaemia/reperfusion models both in dogs and rats (further discussed below; Wang et al., 2010; Nanhwan et al., 2014; Ye et al., 2015). The explanation for the differences between the these studies concerning the cardioprotective effect of clopidogrel is unclear, but might be related to the species, the dose and treatment duration of clopidogrel or the length of reperfusion.

In patients, infarct size was significantly smaller (MRI and enzymic data) and myocardial salvage index was higher, in patients receiving clopidogrel 600 mg compared with 300 mg (Song et al., 2012). Whether this beneficial effect of clopidogrel resulted from direct effects on cardioprotective signalling cascades as suggested by (Yang et al. 2013) or was related to a more potent and quicker platelet inhibition with a 600 mg clopidogrel loading dose remains to be established. In support of the latter hypothesis, clopidogrel improved myocardial blush grade in the 600 mg group more than in the 300 mg group, supporting the notion of improved perfusion and less platelet aggregation (Song et al., 2012).

Ticagrelor

In a dog model of myocardial ischaemia/reperfusion, ticagrelor initiated 5 min before reperfusion reduced infarct size, whereas clopidogrel had no effect despite a similar and almost complete P2Y₁₂ receptor inhibition for clopidogrel and ticagrelor (Wang et al., 2010). Thus, these data indicate a non‐P2Y₁₂ receptor‐mediated cardioprotective effect of ticagrelor. In a rat ischaemia/reperfusion model, ticagrelor but not clopidogrel dose‐dependently reduced infarct size after 7 days p.o. pre‐treatment. The effect was reversed by adenosine receptor blockade providing evidence that the seen cardioprotective effect of ticagrelor was adenosine‐dependent (Nanhwan et al., 2014). In a recent follow‐up study using the same rat model, a single acute administration of ticagrelor just before reperfusion provided a similar reduction in infarct size following ischaemia/reperfusion as previously seen with p.o. pre‐treatment. Again, clopidogrel was ineffective. In addition, infarct size reduction translated into an improved heart function 4 weeks later. Finally, delayed initiation of p.o. ticagrelor treatment starting the day after reperfusion provided a similar improvement in heart function 4 weeks later as that following the single acute treatment (Ye et al., 2015). If these data translate into man, exposure of ticagrelor at the time of PCI could protect against reperfusion injury and improve long‐term heart function. Interestingly, the data indicate that an improved heart function may not exclusively depend on an acute cardioprotective effect as delayed exposure to ticagrelor provided a long‐term benefit. This may be especially relevant for patients with ST‐elevation myocardial infarction (STEMI) that have a delayed increase in the plasma concentration of ticagrelor and thienopyridines (Alexopoulos et al., 2012). In the first rat study, the infarct size reduction by ticagrelor was partly reversed by high‐dose, but not low‐dose, aspirin (Nanhwan et al., 2014) providing a plausible mechanistic explanation for the apparent interaction with high‐dose aspirin in patients receiving ticagrelor seen in the PLATO trial (Mahaffey et al., 2011). Previously, an alternative hypothesis has been proposed related to that the balance between the added anti‐platelet benefit of high‐dose aspirin in combination with P2Y₁₂ receptor antagonism and its potential harm may be different depending on the level of P2Y₁₂ receptor blockade (Kirkby et al., 2011; Björkman et al., 2013; see also Warner et al., 2011). These hypotheses do not contradict each other, but could work together to explain the apparent interaction.

The adenosine‐dependent cardioprotective effect of ticagrelor in animal models may help explain the unique cardiovascular and total mortality benefit seen with ticagrelor in the PLATO trial, but this remains to be shown in man. In the recent PEGASUS study, both 60 and 90 mg of ticagrelor did show a similar beneficial effect on the primary composite endpoint. There was a non‐significant trend for an effect on cardiovascular mortality on both 60 and 90 mg and a numerical decrease in total mortality on 60, but not 90 mg ticagrelor. As the study was not powered to show significant effects on cardiovascular or total mortality, these data need to be interpreted with caution.

The stronger and more rapid platelet inhibition, compared with that after clopidogrel, did not translate into a greater benefit for ticagrelor, in terms of myocardial perfusion in the PLATO angiographic sub‐study (Kunadian et al., 2013). Also, in the ATLANTIC study, when pre‐hospital and in‐hospital ticagrelor administration was compared, there was no difference in terms of myocardial perfusion (Montalescot et al., 2014). Potentially, the short time interval between randomization and PCI, in the PLATO angiographic sub‐study (median 0.68 h), and between pre‐hospital and in‐hospital administration in the ATLANTIC study (median 31 min) was too short to induce any significant difference in the early pharmacodynamic effect between treatments to impact on myocardial perfusion.

Randomized clinical studies comparing the effect of thienopyridines and ticagrelor in terms of infarct size following ischaemia/reperfusion as well as myocardial perfusion are needed. Two such studies are listed on ClinicalTrials.gov. (NCT02233790, NCT02429271) and the design of another has been published (Park et al., 2014). Given the issue of rapid drug absorption specifically in STEMI patients, highlighted previously, the design of such patient studies are challenging.

In summary, experimental evidence support the possibility that thienopyridines and ticagrelor can reduce infarct size following ischaemia/reperfusion. The evidence is especially convincing for ticagrelor as the underlying mechanism, mediated by adenosine, is defined and clear benefit over clopidogrel has been shown in comparative studies.

Conclusion

The effect and clinical benefit of P2Y₁₂ receptor antagonists may not be limited to platelet inhibition and prevention of arterial thrombus formation. Potential additional effects include those secondary to platelet inhibition such as reducing the pro‐inflammatory role of platelets and those that may be related to P2Y₁₂ receptor inhibition on other cells apart from platelets including dyspnea. The thienopyridines and ticagrelor differ in their mode of action on the P2Y₁₂ receptor, and in addition, ticagrelor is unique in having the only well‐documented additional target for inhibition, the adenosine transporter ENT1. These differences provide different opportunities for the P2Y₁₂ receptor antagonists to have effects beyond protection against arterial thrombosis. Explorative studies in humans are needed to translate promising animal data in terms of cardioprotection, vascular function, inflammation and atherosclerosis and to compare the effects of clopidogrel, prasugrel and ticagrelor.

Conflict of interest

Rainer Schulz: research grants from Zealand Pharma; honoraria for lectures and advisory boards from AstraZeneca, Recordati, Sanofi and Servier. Sven Nylander is an employee of AstraZeneca.

Nylander, S. , and Schulz, R. (2016) Effects of P2Y₁₂ receptor antagonists beyond platelet inhibition – comparison of ticagrelor with thienopyridines. British Journal of Pharmacology, 173: 1163–1178. doi: 10.1111/bph.13429.